Grants and Contributions:

Grant or Award spanning more than one fiscal year (2017-2018 to 2018-2019).

Sightline Innovation is a company that provides artificial intelligent products and solutions for businessesx000D
around the world. One area of interest for the company is visual inspection for quality control inx000D
manufacturing. The goal is to detect defects in manufactured parts and classify the type of defects. Since visualx000D
inspection by humans is time-consuming and costly, it is imperative to develop visual inspection systems thatx000D
can automatically detect and classify defects from images. Traditional visual inspection systems are designedx000D
using hand-engineering features. However, since Sightline has clients from multiple industries, the visualx000D
inspection tasks of these clients can be vastly different depending on the industry. It is nearly impossible tox000D
design visual features that can work effectively across various visual inspection tasks. So engineers at Sightlinex000D
end up spending a lot of time hand-tuning the system for each task.x000D
Currently, the company has access to a large amount of images for various visual inspection tasks fromx000D
different clients. The company is interested in exploiting these data to develop machine learning basedx000D
solutions for visual inspection systems that can be quickly adapted to specific clients.x000D
In the proposed research, we will study two specific problems. First, we will develop algorithms for localizingx000D
defects in images from weakly labeled data. Second, we will develop domain adaptation techniques for visualx000D
inspection. This project will produce a set of algorithms and prototypes for visual inspection that the companyx000D
can build upon and integrate with its products.